1 # This file is part of lyx2lyx
2 # Copyright (C) 2002-2011 Dekel Tsur <dekel@lyx.org>,
3 # José Matos <jamatos@lyx.org>, Richard Kimberly Heck <rikiheck@lyx.org>
5 # This program is free software; you can redistribute it and/or
6 # modify it under the terms of the GNU General Public License
7 # as published by the Free Software Foundation; either version 2
8 # of the License, or (at your option) any later version.
10 # This program is distributed in the hope that it will be useful,
11 # but WITHOUT ANY WARRANTY; without even the implied warranty of
12 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 # GNU General Public License for more details.
15 # You should have received a copy of the GNU General Public License
16 # along with this program; if not, write to the Free Software
17 # Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
21 This module offers several free functions to help parse lines.
22 More documentaton is below, but here is a quick guide to what
23 they do. Optional arguments are marked by brackets.
25 find_token(lines, token[, start[, end[, ignorews]]]):
26 Returns the first line i, start <= i < end, on which
27 token is found at the beginning. Returns -1 if not
29 If ignorews is (given and) True, then differences
30 in whitespace do not count, except that there must be no
31 extra whitespace following token itself.
33 find_token_exact(lines, token[, start[, end]]]):
34 As find_token, but with ignorews set to True.
36 find_tokens(lines, tokens[, start[, end[, ignorews]]]):
37 Returns the first line i, start <= i < end, on which
38 one of the tokens in tokens is found at the beginning.
39 Returns -1 if not found.
40 If ignorews is (given and) True, then differences
41 in whitespace do not count, except that there must be no
42 extra whitespace following token itself.
44 find_tokens_exact(lines, token[, start[, end]]):
45 As find_tokens, but with ignorews True.
47 find_token_backwards(lines, token, start):
48 find_tokens_backwards(lines, tokens, start):
49 As before, but look backwards.
51 find_substring(lines, sub[, start[, end]]) -> int
52 As find_token, but sub may be anywhere in the line.
54 find_re(lines, rexp, start[, end]):
55 As find_token, but rexp is a regular expression object,
56 so it has to be passed as e.g.: re.compile(r'...').
58 get_value(lines, token[, start[, end[, default[, delete]]]]):
59 Similar to find_token, but it returns what follows the
60 token on the found line. Example:
61 get_value(document.header, "\\use_xetex", 0)
62 will find a line like:
64 and, in that case, return "true". (Note that whitespace
65 is stripped.) The final argument, default, defaults to "",
66 and is what is returned if we do not find anything. So you
67 can use that to set a default.
68 If delete is True, then delete the line if found.
70 get_quoted_value(lines, token[, start[, end[, default[, delete]]]]):
71 Similar to get_value, but it will strip quotes off the
72 value, if they are present. So use this one for cases
73 where the value is normally quoted.
75 get_option_value(line, option):
76 This assumes we have a line with something like:
78 and returns value. Returns "" if not found.
80 get_bool_value(lines, token[, start[, end[, default, delete]]]]):
81 Like get_value, but returns a boolean.
83 set_bool_value(lines, token, value[, start[, end]]):
84 Find `token` in `lines[start:end]` and set to boolean value bool(`value`).
85 Return old value. Raise ValueError if token is not in lines.
87 del_token(lines, token[, start[, end]]):
88 Like find_token, but deletes the line if it finds one.
89 Returns True if a line got deleted, otherwise False.
91 Use get_* with the optional argument "delete=True", if you want to
92 get and delete a token.
94 find_beginning_of(lines, i, start_token, end_token):
95 Here, start_token and end_token are meant to be a matching
96 pair, like "\\begin_layout" and "\\end_layout". We look for
97 the start_token that pairs with the end_token that occurs
98 on or after line i. Returns -1 if not found.
99 So, in the layout case, this would find the \\begin_layout
100 for the layout line i is in.
102 ec = find_token(document.body, "</cell", i)
103 bc = find_beginning_of(document.body, ec, \
105 Now, assuming no -1s, bc-ec wraps the cell for line i.
107 find_end_of(lines, i, start_token, end_token):
108 Like find_beginning_of, but looking for the matching
109 end_token. This might look like:
110 bc = find_token_(document.body, "<cell", i)
111 ec = find_end_of(document.body, bc, "<cell", "</cell")
112 Now, assuming no -1s, bc-ec wrap the next cell.
114 find_end_of_inset(lines, i):
115 Specialization of find_end_of for insets.
117 find_end_of_layout(lines, i):
118 Specialization of find_end_of for layouts.
120 find_end_of_sequence(lines, i):
121 Find the end of the sequence of layouts of the same kind.
122 Considers nesting. If the last paragraph in sequence is nested,
123 the position of the last \\end_deeper is returned, else
124 the position of the last \\end_layout.
126 is_in_inset(lines, i, inset, default=(-1,-1)):
127 Check if line i is in an inset of the given type.
128 If so, returns starting and ending lines. Otherwise,
131 is_in_inset(document.body, i, "\\begin_inset Tabular")
132 returns (-1,-1) unless i is within a table. If it is, then
133 it returns the line on which the table begins and the one
134 on which it ends. Note that this pair will evaulate to
136 if is_in_inset(..., default=False):
137 will do what you expect.
139 get_containing_inset(lines, i):
140 Finds out what kind of inset line i is within. Returns a
141 list containing what follows \\begin_inset on the line
142 on which the inset begins, plus the starting and ending line.
143 Returns False on any kind of error or if it isn't in an inset.
144 So get_containing_inset(document.body, i) might return:
145 ("CommandInset ref", 300, 306)
146 if i is within an InsetRef beginning on line 300 and ending
149 get_containing_layout(lines, i):
150 As get_containing_inset, but for layout. Additionally returns the
151 position of real paragraph start (after par params) as 4th value.
153 find_nonempty_line(lines, start[, end):
154 Finds the next non-empty line.
156 check_token(line, token):
157 Does line begin with token?
159 is_nonempty_line(line):
160 Does line contain something besides whitespace?
162 count_pars_in_inset(lines, i):
163 Counts the paragraphs inside an inset.
170 # Utilities for one line
171 def check_token(line, token):
172 """check_token(line, token) -> bool
174 Return True if token is present in line and is the first element
177 Deprecated. Use line.startswith(token).
179 return line.startswith(token)
182 def is_nonempty_line(line):
183 """is_nonempty_line(line) -> bool
185 Return False if line is either empty or it has only whitespaces,
187 return bool(line.strip())
190 # Utilities for a list of lines
191 def find_token(lines, token, start=0, end=0, ignorews=False):
192 """find_token(lines, token, start[[, end], ignorews]) -> int
194 Return the lowest line where token is found, and is the first
195 element, in lines[start, end].
197 If ignorews is True (default is False), then differences in
198 whitespace are ignored, but there must be whitespace following
201 Use find_substring(lines, sub) to find a substring anywhere in `lines`.
203 Return -1 on failure."""
205 if end == 0 or end > len(lines):
209 for i in range(start, end):
217 if lines[i].startswith(token):
222 def find_token_exact(lines, token, start=0, end=0):
223 return find_token(lines, token, start, end, True)
226 def find_tokens(lines, tokens, start=0, end=0, ignorews=False):
227 """find_tokens(lines, tokens, start[[, end], ignorews]) -> int
229 Return the lowest line where one token in tokens is found, and is
230 the first element, in lines[start, end].
232 Return -1 on failure."""
234 if end == 0 or end > len(lines):
237 for i in range(start, end):
247 if lines[i].startswith(token):
252 def find_tokens_exact(lines, tokens, start=0, end=0):
253 return find_tokens(lines, tokens, start, end, True)
256 def find_substring(lines, sub, start=0, end=0):
257 """find_substring(lines, sub[, start[, end]]) -> int
259 Return the lowest line number `i` in [start, end] where
260 `sub` is a substring of line[i].
262 Return -1 on failure."""
264 if end == 0 or end > len(lines):
266 for i in range(start, end):
272 def find_re(lines, rexp, start=0, end=0):
273 """find_re(lines, rexp[, start[, end]]) -> int
275 Return the lowest line number `i` in [start, end] where the regular
276 expression object `rexp` matches at the beginning of line[i].
277 Return -1 on failure.
279 Start your pattern with the wildcard ".*" to find a match anywhere in a
280 line. Use find_substring() to find a substring anywhere in the lines.
282 if end == 0 or end > len(lines):
284 for i in range(start, end):
285 if rexp.match(lines[i]):
290 def find_token_backwards(lines, token, start):
291 """find_token_backwards(lines, token, start) -> int
293 Return the highest line where token is found, and is the first
294 element, in lines[start, end].
296 Return -1 on failure."""
297 for i in range(start, -1, -1):
298 if lines[i].startswith(token):
303 def find_tokens_backwards(lines, tokens, start):
304 """find_tokens_backwards(lines, token, start) -> int
306 Return the highest line where token is found, and is the first
307 element, in lines[end, start].
309 Return -1 on failure."""
310 for i in range(start, -1, -1):
313 if line.startswith(token):
318 def find_complete_lines(lines, sublines, start=0, end=0):
319 """Find first occurence of sequence `sublines` in list `lines`.
320 Return index of first line or -1 on failure.
322 Efficient search for a sub-list in a large list. Works for any values.
324 >>> find_complete_lines([1, 2, 3, 1, 1, 2], [1, 2])
327 The `start` and `end` arguments work similar to list.index()
329 >>> find_complete_lines([1, 2, 3, 1, 1, 2], [1, 2], start=1)
331 >>> find_complete_lines([1, 2, 3, 1, 1, 2], [1, 2], start=1, end=4)
334 The return value can be used to substitute the sub-list.
335 Take care to check before use:
338 >>> s = find_complete_lines(l, [1, 2])
340 ... l[s : s + 2] = [3]
344 See also del_complete_lines().
348 end = end or len(lines)
352 for j, value in enumerate(sublines):
353 i = lines.index(value, start, end)
360 except ValueError: # `sublines` not found
364 def find_across_lines(lines, sub, start=0, end=0):
365 sublines = sub.splitlines()
366 if len(sublines) > 2:
367 # at least 3 lines: the middle one(s) are complete -> use index search
368 i = find_complete_lines(lines, sublines[1:-1], start + 1, end - 1)
372 if lines[i - 1].endswith(sublines[0]) and lines[i + len(sublines)].startswith(
378 elif len(sublines) > 1:
379 # last subline must start a line
380 i = find_token(lines, sublines[-1], start, end)
383 if lines[i - 1].endswith(sublines[0]):
385 else: # no line-break, may be in the middle of a line
386 if end == 0 or end > len(lines):
388 for i in range(start, end):
394 def get_value(lines, token, start=0, end=0, default="", delete=False):
395 """Find `token` in `lines` and return part of line that follows it.
397 Find the next line that looks like:
398 token followed by other stuff
400 If `delete` is True, delete the line (if found).
402 Return "followed by other stuff" with leading and trailing
405 i = find_token_exact(lines, token, start, end)
408 # TODO: establish desired behaviour, eventually change to
409 # return lines.pop(i)[len(token):].strip() # or default
410 # see test_parser_tools.py
411 l = lines[i].split(None, 1)
419 def get_quoted_value(lines, token, start=0, end=0, default="", delete=False):
420 """get_quoted_value(lines, token, start[[, end], default]) -> string
422 Find the next line that looks like:
423 token "followed by other stuff"
424 Returns "followed by other stuff" with leading and trailing
425 whitespace and quotes removed. If there are no quotes, that is OK too.
426 So use get_value to preserve possible quotes, this one to remove them,
428 Note that we will NOT strip quotes from default!
430 val = get_value(lines, token, start, end, "", delete)
433 return val.strip('"')
436 bool_values = {"true": True, "1": True, "false": False, "0": False}
439 def get_bool_value(lines, token, start=0, end=0, default=None, delete=False):
440 """get_bool_value(lines, token, start[[, end], default]) -> string
442 Find the next line that looks like:
445 Return True if <bool_value> is 1 or "true", False if <bool_value>
446 is 0 or "false", else `default`.
448 val = get_quoted_value(lines, token, start, end, default, delete)
449 return bool_values.get(val, default)
452 def set_bool_value(lines, token, value, start=0, end=0):
453 """Find `token` in `lines` and set to bool(`value`).
455 Return previous value. Raise `ValueError` if `token` is not in lines.
457 Cf. find_token(), get_bool_value().
459 i = find_token(lines, token, start, end)
462 oldvalue = get_bool_value(lines, token, i, i + 1)
463 if oldvalue is value:
466 if get_quoted_value(lines, token, i, i + 1) in ("0", "1"):
467 lines[i] = "%s %d" % (token, value)
469 lines[i] = f"{token} {str(value).lower()}"
474 def get_option_value(line, option):
475 rx = option + r'\s*=\s*"([^"]+)"'
483 def set_option_value(line, option, value):
484 rx = "(" + option + r'\s*=\s*")[^"]+"'
489 return re.sub(rx, r"\g<1>" + value + '"', line)
492 def del_token(lines, token, start=0, end=0):
493 """del_token(lines, token, start, end) -> int
495 Find the first line in lines where token is the first element
496 and delete that line. Returns True if we deleted a line, False
499 k = find_token_exact(lines, token, start, end)
506 def del_complete_lines(lines, sublines, start=0, end=0):
507 """Delete first occurence of `sublines` in list `lines`.
509 Efficient deletion of a sub-list in a list. Works for any values.
510 The `start` and `end` arguments work similar to list.index()
512 Returns True if a deletion was done and False if not.
514 >>> l = [1, 0, 1, 1, 1, 2]
515 >>> del_complete_lines(l, [0, 1, 1])
520 i = find_complete_lines(lines, sublines, start, end)
523 del lines[i : i + len(sublines)]
527 def del_value(lines, token, start=0, end=0, default=None):
529 Find the next line that looks like:
530 token followed by other stuff
531 Delete that line and return "followed by other stuff"
532 with leading and trailing whitespace removed.
534 If token is not found, return `default`.
536 i = find_token_exact(lines, token, start, end)
539 return lines.pop(i)[len(token) :].strip()
542 def find_beginning_of(lines, i, start_token, end_token):
545 i = find_tokens_backwards(lines, [start_token, end_token], i - 1)
548 if lines[i].startswith(end_token):
557 def find_end_of(lines, i, start_token, end_token):
561 i = find_tokens(lines, [end_token, start_token], i + 1)
564 if lines[i].startswith(start_token):
573 def find_nonempty_line(lines, start=0, end=0):
576 for i in range(start, end):
582 def find_end_of_inset(lines, i):
583 "Find end of inset, where lines[i] is included."
584 return find_end_of(lines, i, "\\begin_inset", "\\end_inset")
587 def find_end_of_layout(lines, i):
588 "Find end of layout, where lines[i] is included."
589 return find_end_of(lines, i, "\\begin_layout", "\\end_layout")
592 def is_in_inset(lines, i, inset, default=(-1, -1)):
594 Check if line i is in an inset of the given type.
595 If so, return starting and ending lines, otherwise `default`.
597 is_in_inset(document.body, i, "\\begin_inset Tabular")
598 returns (-1,-1) if `i` is not within a "Tabular" inset (i.e. a table).
599 If it is, then it returns the line on which the table begins and the one
601 Note that this pair will evaulate to boolean True, so (with the optional
602 default value set to False)
603 if is_in_inset(..., default=False):
604 will do what you expect.
606 start = find_token_backwards(lines, inset, i)
609 end = find_end_of_inset(lines, start)
610 if end < i: # this includes the notfound case.
615 def get_containing_inset(lines, i):
617 Finds out what kind of inset line i is within. Returns a
618 list containing (i) what follows \\begin_inset on the line
619 on which the inset begins, plus the starting and ending line.
620 Returns False on any kind of error or if it isn't in an inset.
624 stins = find_token_backwards(lines, "\\begin_inset", j)
627 endins = find_end_of_inset(lines, stins)
635 inset = get_value(lines, "\\begin_inset", stins)
639 return (inset, stins, endins)
642 def get_containing_layout(lines, i):
644 Find out what kind of layout line `i` is within.
646 (layoutname, layoutstart, layoutend, startofcontent)
650 * end line number, and
651 * number of first paragraph line (after all params).
652 Return `False` on any kind of error.
656 stlay = find_token_backwards(lines, "\\begin_layout", j)
659 endlay = find_end_of_layout(lines, stlay)
667 layoutname = get_value(lines, "\\begin_layout", stlay)
668 if layoutname == "": # layout style missing
669 # TODO: What shall we do in this case?
671 # layoutname == "Standard" # use same fallback as the LyX parser:
672 # raise ValueError("Missing layout name on line %d"%stlay) # diagnosis
673 # return False # generic error response
679 "\\start_of_appendix",
680 "\\paragraph_spacing",
682 "\\labelwidthstring",
687 if lines[stpar].split(" ", 1)[0] not in par_params:
689 return (layoutname, stlay, endlay, stpar)
692 def count_pars_in_inset(lines, i):
694 Counts the paragraphs within this inset
696 ins = get_containing_inset(lines, i)
700 for j in range(ins[1], ins[2]):
701 m = re.match(r"\\begin_layout (.*)", lines[j])
702 found_inset = get_containing_inset(lines, j)
703 if m and found_inset and found_inset[1] == ins[1]:
709 def find_end_of_sequence(lines, i):
711 Returns the end of a sequence of identical layouts.
713 lay = get_containing_layout(lines, i)
720 m = re.match(r"\\begin_layout (.*)", lines[i])
721 if m and m.group(1) != layout:
723 elif lines[i] == "\\begin_deeper":
724 j = find_end_of(lines, i, "\\begin_deeper", "\\end_deeper")
729 if m and m.group(1) == layout:
730 endlay = find_end_of_layout(lines, i)
733 if i == len(lines) - 1: